plans for long term support releases?

[Thomas M Steenholdt]

Arjan van de Ven wrote:
> On Tue, 2007-01-16 at 07:04 +0100, Thomas M Steenholdt wrote:
>> Now that Core and Extras are going to be merged and the distro is 
>> opening up to become even more (?) community driven, has anyone played 
>> with the though of eventually releasing a long term support version of 
>> Fedora?
>>
>> It could be a based on a staple snapshot of Fedora 7 + 4 months worth of 
>> updates or whatever, at this point I'm more interested in hearing about 
>> the idea than the details, which will surely follow, if i'm not the only 
>> one who think this could be a good idea. Especially now that we're going 
>> to do special server spins etc...
>>
>> Just a thought (hope this was not brought up ages ago and I just missed it)
> 
> 
> so one question: how many packages are YOU willing to be the person for
> that keeps up with the security stuff? (it should be packages you're
> familiar with since security backports are generally quite hard, and
> incrementally harder the older the package)....
> 

In my mind, something like this could be accomplished in several ways. 
If backporting is too much work (I know it takes time) then perhaps it 
could be done by following upstream releases like we do today, but for a 
longer time. LTS in this case is meant to decrease the burden of having 
to provide updates 3-5 years ahead for every version of Fedora. We could 
provide a Fedora LTS, that's essentially a normal Fedora that we choose 
to provide updates for, for a good while longer that all the other Fedoras.

Perhaps the solution is to re-launch legacy in a more integrated way?!? 
I heard the cut back, but I don't know how much.

But hey, these are only thoughts. Perhaps the official answer could be 
"for long terms support cases, go with RHEL/CentOS"?

> if the answer is "eh none" then you are giving the same answer a LOT of
> people do. THere's nothing wrong with that per se, but there IS
> something wrong with the "but I want other people to do it for me. For
> free. Without even saying thanks". 
> 
> Fedora Legacy may have had it's faults, but there is also a fundamental
> contradiction in it.
> 
> A lot of people like fedora for it's latest stuff. And there's people
> who just want a distro stable and keep using it.
> 
> The first group generally moves to a new FC release within 3 months
> after it being released. Fair enough, and this is actually quite a
> really large group. Even in the RHL days (with 2+ years of support), 90%
> of the users moved to a new release within that timeframe. 

Absolutely - And I do that too. Within the first month I'd say. But 
that's not always possible/feasible for server systems.

> 
> The other 10% wants things to "just work", and this is pretty much the
> exact same crowd who only want stable tested packages. So.. your testing
> base for updates is... exactly zero. (This problem isn't unique to
> fedora. For example the 2.4 kernel series has exactly this same problem;
> nobody who still uses 2.4 wants to test prereleases)
> 
> What is worse, usually developers (either upstream coders or package
> maintainers) do not fall in this 10%, they are more in the 90% by *far*.
> And for the part they're in the 10%, they don't want those systems
> unstable/testing either...
> 
> My conclusion on this (and I've been outside a distro for a while now,
> looking at the entire ecosystem from a distance) is that you either
> 1) have people paid to do the work
> or
> 2) have longer release cycles (with one in development and one in
> maintenance) 
> 
> debian follows the 2) model.. RHEL, SLES and Ubuntu the 1) model.
> Thing is... if I ask YOU the question, how many dollars are you willing
> to pay for 6 extra months of maintenance, I can guess the answer with a
> dollar or two :)
> 
> Note that this includes QA; without QA this entire thing is worthless,
> since if update packages break more than once for the 10% that cares,
> they will yell, scream and claim the long term support effectively is
> worthless and doesn't exist anyway.
> 

I agree completely. Perhaps an LTS is simply out of scope for Fedora and 
that's okay. I can easily use RHEL/CentOS, I just though it was a neat 
idea if we could think of a way that would make it possible to install a 
special Fedora spin that had the property of LTS. nothing more. Of 
course things have to work in sense of people, money etc.

Perhaps setup a Fedora donation thingie, where happy users can 
contribute and see what kind of interesting prospects something like 
that could bring (if any).

/Thomas